REVIEWS
Microbial biogeography: putting
microorganisms on the map
Jennifer B. Hughes Martiny*, Brendan J.M. Bohannan‡, James H. Brown§,
Robert K. Colwell ||, Jed A. Fuhrman¶, Jessica L. Green**, M. Claire Horner-Devine‡‡,
Matthew Kane§§, Jennifer Adams Krumins|| || , Cheryl R. Kuske¶¶, Peter J. Morin|| || ,
Shahid Naeem***, Lise Øvreås‡‡‡, Anna-Louise Reysenbach§§§, Val H. Smith|| || ||
and James T. Staley¶¶¶
Abstract | We review the biogeography of microorganisms in light of the biogeography of
macroorganisms. A large body of research supports the idea that free-living microbial taxa
exhibit biogeographic patterns. Current evidence confirms that, as proposed by the
Baas-Becking hypothesis, ‘the environment selects’ and is, in part, responsible for spatial
variation in microbial diversity. However, recent studies also dispute the idea that
‘everything is everywhere’. We also consider how the processes that generate and maintain
biogeographic patterns in macroorganisms could operate in the microbial world.
*Department of Ecology
and Evolutionary Biology,
80 Waterman Street,
BOX G-W, Brown University,
Providence, Rhode Island
02912, USA.
Correspondence to J.B.H.M.
e-mail: jennifer_martiny
@brown.edu
doi:10.1038/nrmicro1341
Biogeography is the study of the distribution of biodiversity
over space and time. It aims to reveal where organisms
live, at what abundance, and why. The study of biogeography offers insights into the mechanisms that generate
and maintain diversity, such as speciation, extinction,
dispersal and species interactions1. Since the eighteenth
century, biologists have investigated the geographic distribution of plant and animal diversity. More recently, the
geographic distributions of microorganisms have been
examined. Genetic methodologies have revealed that past
culture-based studies missed most microbial diversity2–4,
and have allowed recent studies to sample microbial diversity more deeply and widely than ever before5,6. Microbial
biogeography stands to benefit tremendously from these
advances, although there is still debate as to whether
microorganisms exhibit any biogeographic patterns7–10.
Although traditionally confined to separate academic
disciplines, ecologists who study microorganisms and
those who study macroorganisms have been interacting
more often in recent years. Indeed, this article is a result
of a National Center for Ecological Analysis and Synthesis
(NCEAS) working group composed of scientists from
both specialties. Our goal here is to review what is known
regarding the biogeography of microorganisms in light
of that of macroorganisms. This inquiry is not driven by
the expectation that microorganisms simply follow the
patterns of macroorganisms, but rather by the fact that
the biogeography of macroorganisms is much better
studied. Furthermore, micro- and macroorganisms are
often involved in intimate associations that affect each
other’s geographic distributions11,12. Therefore, a logical
first hypothesis is that the biogeography of microorganisms is similar to the biogeography of macroorganisms.
To the extent that microorganisms conform to the relationships documented for macroorganisms, they will
extend the generality of empirical patterns and support
mechanistic hypotheses that all living entities share universal attributes. Alternatively, if microorganisms can be
shown to represent clear exceptions to the biogeographic
patterns of plants and animals, then this will call attention
to unique features of microbial life that have influenced
the generation and maintenance of its diversity.
As an initial step towards distinguishing between
these hypotheses, we suggest a framework for investigating whether microbial assemblages differ in
different places and the extent to which this spatial
variation is due to contemporary environmental factors and historical contingencies. We then discuss the
mechanistic processes that generate and maintain biogeographic patterns in macroorganisms and consider
their relevance to microbial biogeography.
There is no universal definition of a ‘microorganism’.
The term generally denotes members of the domains
Bacteria and Archaea, as well as microscopic members of the domain Eukarya (for example, unicellular
algae, some fungi and protists). For convenience, we
further define a microorganism as having a mass
of less than 10–5 g and a length of less than 500 µm.
102 | FEBRUARY 2006 | VOLUME 4
www.nature.com/reviews/micro
© 2006 Nature Publishing Group
REVIEWS
Province
A region the biotic composition
of which reflects the legacies of
historical events.
Habitat type
An environment defined by the
suite of its abiotic and biotic
characteristics.
We do not consider the question of whether viruses have
biogeography, as their biology adds further complications
and, in most cases, far less is known about their distribution than that of other microorganisms (for a recent
discussion see REF. 13).
A framework for microbial biogeography
A long-standing theme of traditional biogeography
is the relative influence of contemporary environmental factors versus the legacies of historical events
on present-day distribution patterns. In the early
nineteenth century, Augustin P. de Candolle 14 distinguished between the influence of ‘habitations’ and
‘stations’ on the distribution of plant diversity. He used
the word ‘habitation’ to signify a biotic province. For
instance, the many plant and animal species unique
to Australia are attributable to past connections to,
and long isolation from, other continents, and clearly
distinguish Australia as a distinct province. Augustin
P. de Candolle used the word ‘station’ to mean a habitat
type, or a constellation of contemporary abiotic and
biotic environmental variables that influenced plant
composition. For instance, Australia contains various
habitat types that support different biotic assemblages;
some habitat types are unique to the province (such as
the Mallee scrublands), whereas others are found in
many provinces (such as the coastal scrub habitat that
has analogues in California, Chile, South Africa and
the Mediterranean).
Continents are especially clear examples of macroorganism provinces, but we use the term ‘province’ more
freely than plant and animal provinces have been traditionally defined. A province is any area, the biota of which
reflects historical events. Therefore, province size might
vary greatly and depend on the particular taxon and
resolution of focus. For instance, two lakes a hundred kilometers apart might be separate provinces for a particular
strain of bacteria, but all the lakes on a continent might be
part of the same province for protist assemblages.
Author addresses
‡
Department of Biological Sciences, Stanford University, California 94305, USA.
Department of Biology, University of New Mexico, New Mexico 87131, USA.
||
Department of Ecology and Evolutionary Biology, University of Connecticut,
Connecticut 06269, USA.
¶
Department of Biological Sciences, University of Southern California,
California 90089, USA.
**School of Natural Sciences, University of California, Merced, California 95344, USA.
‡‡
School of Aquatic and Fishery Science, University of Washington,
Washington 98195, USA.
§§
Division of Molecular and Cellular Biosciences, National Science Foundation,
Virginia 22230, USA.
|| ||
Department of Ecology and Evolution, Rutgers University, New Jersey 08901, USA.
¶¶
Bioscience Division, Los Alamos National Laboratory, New Mexico 87545, USA.
***Department of Ecology, Evolution and Environmental Biology, Columbia
University, New York 10027, USA.
‡‡‡
Department of Biology, University of Bergen, Norway.
§§§
Department of Biology, Portland State University, Oregon 97207, USA.
||||||
Department of Ecology and Evolutionary Biology, University of Kansas,
Kansas 66045, USA.
¶¶¶
Department of Microbiology, University of Washington, Washington 98195, USA.
§
The consideration of habitats and provinces provides a useful framework for addressing four alternative
hypotheses. The null hypothesis is that microorganisms
are randomly distributed over space. In this case,
microorganisms essentially experience only one habitat and one province. A second hypothesis is that the
biogeography of microorganisms reflects the influence
of contemporary environmental variation (multiple
habitats) within a single province. This is the so-called
Baas-Becking hypothesis — for microbial taxa, ‘everything is everywhere — the environment selects’15,16.
The claim that ‘the environment selects’ implies that
different contemporary environments maintain distinctive microbial assemblages. The claim ‘everything
is everywhere’ implies that microorganisms have such
enormous dispersal capabilities that they rapidly erase
the effects of past evolutionary and ecological events.
A third alternative is that all spatial variation is due to the
lingering effects of historical events (multiple provinces
but only one habitat). Historical events that might influence present-day assemblages include dispersal limitation and past environmental conditions, both of which
can lead to genetic divergence of microbial assemblages.
We consider dispersal limitation to be a historical event,
as current composition is influenced by past dispersal
limitation, whether relatively recent or ancient. The final
hypothesis is that the distributions of microbial taxa, like
those of macroorganisms, reflect the influences of both
past events and contemporary environmental conditions
— in other words, that microbial distributions are shaped
by multiple habitats and multiple provinces.
Distinguishing between the four hypotheses
addresses two central biogeography questions: first,
do microbial assemblages differ in different locations
(do microorganisms have biogeography); and second,
if microbial assemblages do differ by location, is the
spatial variation due to present-day environmental factors, historical contingencies, or both? By definition,
differences in microbial assemblages are due to variation in the relative abundances of taxa, including the
presence of a particular taxon in one assemblage and its
absence in another. As such, we focus on how the relative abundances of microbial taxa vary over space, rather
than whether any microbial taxa are truly restricted to
particular geographic areas, as it is nearly impossible
to conclusively show that a microbial taxon is absent
from a given location.
Do microorganisms have biogeography?
It has long been known that many host-associated
microorganisms exhibit patterns of genetic, morphological and functional differentiation that are related to
the distribution of their hosts10,12,17–19. Now, a growing
body of evidence shows that free-living microorganisms
also vary in abundance, distribution and diversity, over
various taxonomic and spatial scales (some examples are
given in TABLE 1).
The simplest demonstration of microbial biogeography
is that microbial composition across a landscape is nonrandom, thereby rejecting the first hypothesis above. For
example, Cho and Tiedje20 showed that genetic distance
NATURE REVIEWS | MICROBIOLOGY
VOLUME 4 | FEBRUARY 2006 | 103
© 2006 Nature Publishing Group
REVIEWS
Table 1 | Examples of studies that have found non-random distributions of free-living microbial taxa
Organisms
Approximate
scale (km)
Habitat
Unit
Correlated with
Linear distance
Ref.
Pseudomonads*
20,000
Soil
BOX-PCR isolation
3-CBD bacteria‡
20,000
Soil
ARDRA isolation
20
Aerobic, anoxygenic phototrophs‡
20,000
Marine
Dissociation curves
Latitude
91
SAR11 bacteria and archaea‡
13,000
Marine
16S/ITS sequence
Depth
92
Green sulphur bacteria‡
8,000
Lakes
16S sequence
Continental divide
93
N-fixing bacteria‡
700
Desert crusts
Sequence and TRFLP of nifH
and 16S
Mature versus poorly
developed crusts
94
Crenarchaeota*
200
Soil
PCR-SSCP of 16S
At small scales, distance
95
Crenarchaeota*
200
Soil
PCR-SSCP of 16S
Rhizosphere versus bulk soil
96
90
Bacteria‡
50
Marine
DGGE of 16S
Ocean front
97
Bacteria*
35
Marine
DGGE of 16S
Depth and ocean front
98
‡
Bacteria
15
River plume
DGGE of 16S
River–marine transition
99
Bacteria*
5
River plume
DGGE of 16S
Salinity
22
Bacteria, archaea and eukaryotes*
3
Salterns
DGGE, TRFLP, RISA
Salinity
100
Pseudomonas cepacia*
3
Soil
Isolate allozymes
Vegetation
101
Bacteria and eukaryotes*
1
Soil
RNA hybridization
Cultivation history
102
Gram-negative bacteria*
0.8
Soil
sole carbon source
Latitude
103
Microorganisms*
0.2
Groundwater
RAPD
Oxygen zonation
Microorganisms*
0.1
Agricultural soil
AFLP
Bacteria and archaea‡
0.02
Lake
DGGE of 16S
Depth
Bacteria*
0.01
Drinking water
TRFLP of 16S
Bulkwater versus pipe
biofilm
Purple non-sulphur bacteria‡
0.01
Fresh marsh
BOX-PCR isolation
Linear distance
Bacteria*
0.01
Soil
RFLPs of 16S
Microorganisms*
0.002
Salt marsh
RAPD
104
105
23
106
21
28
Marsh elevation
107
The studies are ordered by the geographic scale over which the samples were taken, reported as the approximate furthest distance between sampling points.
*The study found significant non-random distributions. ‡No statistics were performed. If the authors reported that the pattern in microbial composition was correlated
with an environmental characteristic, this is reported in the ‘correlated with’ column, even if this relationship was not statistically tested. 3-CBD, 3-chlorobenzoatedegrading; AFLP, amplified fragment length polymorphism; ARDRA, amplified ribosomal DNA restriction analysis; DGGE, denaturing gradient gel electrophoresis;
ITS, intergenic transcribed space; nifH, bacterial gene that encodes for nitrogenase; PCR-SSCP, polymerase chain reaction-single strand conformational polymorphism;
RAPD, random amplified polymorphic DNA; RFLP, restriction fragment length polymorphisms; RISA, ribosomal intergenic spacer analysis; TRFLP, terminal RFLP.
Beta diversity
Taxonomic diversity due to
turnover in composition
between assemblages.
between fluorescent pseudomonads was related to geographic distance. Similarly, Oda et al.21 showed genetic
differences among purple non-sulphur bacteria along a
10-meter marsh transect. Many of the studies listed in
TABLE 1 find correlations between assemblage composition and environmental or geographic characteristics,
such as salinity22, depth23 and latitude24.
Taxa–area relationships are further evidence for
microbial biogeography. An increase in the number
of taxa observed with increasing sample area (often
referred to as a species–area relationship) has been
detected repeatedly in plants and animals25. Recently,
investigators have reported similar patterns in microbial communities, in both contiguous26–28 and island29,30
habitats. Within contiguous habitats, a positive taxa–area
relationship might arise even if microorganisms are randomly distributed over space31,32. By contrast, taxa–area
relationships that reflect increasing spatial heterogeneity
of biotic composition (beta diversity) at increasing spatial
scales will exhibit a decrease in biotic similarity with
spatial separation — a striking, non-random pattern.
Distinguishing between environment and history
A limitation of the analyses listed in TABLE 1 is that they
exclude only the hypothesis that microbial assemblages
are spatially random. This leaves the problem of determining how much of the spatial variation in microbial
distributions and assemblages is due to contemporary
environmental conditions or historical contingencies.
Answering this question requires information on the
current abiotic and biotic conditions and the spatial
arrangement of the sampled assemblages.
Consider first the case of sampling discrete, pre-defined
habitat types that might or might not influence microbial
composition. These habitats might be different depths in
the ocean water column or rhizospheres of different plant
species. As a simple example, consider the case of two distinct geographic locations, each containing three discrete
habitat types from which two replicate samples are taken,
for a total of 12 sampling sites. The microbial assemblage
of each sample is analysed using methods such as clonelibrary sequence analysis, a community fingerprinting
technique or culturing techniques.
104 | FEBRUARY 2006 | VOLUME 4
www.nature.com/reviews/micro
© 2006 Nature Publishing Group
REVIEWS
Box 1 | Resemblance matrices for biogeographic analyses
Abundance or
Three square resemblance (similarity or distance) matrices
incidence matrix
Biotic-similarity
are fundamental for biogeographic analyses35. These
matrix
1 2 3 4 5
matrices are derived from three data matrices (see figure).
a
1
2 3 4 5
The presence/absence or abundance data from all the sites
b
1
are first summarized in an incidence or abundance matrix,
c
2
d
in which letters are the taxa and numbers are the sites in the
3
e
4
figure. The taxa are defined by any appropriate operational
f
5
taxonomic unit (OTU), such as a sequence-similarity cut-off
g
or fingerprint-band length. To calculate the biotic-similarity
h
matrix, the composition is compared between each pair of
Environmental
sites and a similarity index is calculated. The similarity index
matrix
Environmental-similarity
might be based on presence and absence of each taxon, such
matrix
1 2 3 4 5
as the classic Jaccard index, or also incorporate abundance,
i
1 2 3 4 5
such as the Morisita–Horn index84 or Chao’s abundanceii
1
based Jaccard estimator85. The diagonal entries of the
iii
2
iv
3
similarity matrix are ‘1’s, and the values above and below the
v
4
diagonal are mirror images.
vi
5
The environmental matrix reports the values of each
vii
environmental parameter (Roman numerals) recorded at
Geographic-distance
each site and is transformed into an environmental-similarity
Site-location
matrix
matrix. One possible similarity index to use is 1 minus a
matrix
1
2 3 4 5
standardized Euclidean distance, in which raw values of
1 2 3 4 5
1
x
environmental variables are first transformed to their
2
y
standard normal deviate equivalents ([x – mean] divided by
3
4
the standard deviation) to accommodate the different units
5
of the different variables27. In the absence of prior knowledge
of which variables influence the microbial community of
interest, a large number of factors are often measured. In this case, preliminary analyses are useful to determine the
variables that relate to community composition; adding in many unrelated variables can swamp out the signature of any
significant variables.
The third matrix is a geographic-distance matrix and is usually the actual geographic distances between each
pair of sites, which can be calculated from latitude and longitude values (X and Y). The diagonal values are zero.
In some cases, one might want to weight the cell values of a geographic-distance matrix to account for potential
barriers to dispersal86. For instance, one could account for ocean currents or land masses when investigating
marine communities.
Distance effect
The influence of isolation on
biotic composition after
controlling for the influence of
the contemporary
environment.
Genetic drift
Changes in gene frequencies in
a population caused solely by
chance.
The similarities between each sampling assemblage
can be summarized in a biotic-similarity matrix (BOX 1).
To picture these data, this matrix can be collapsed with a
clustering algorithm. The results of this analysis can then
be displayed as a dendrogram33 or along dimensionless
axes with multidimensional scaling34. To test the four
alternative hypotheses, one can then overlay the information on habitat types and geographic location on the
assemblage clustering. FIGURE 1a illustrates an example.
The green and white circles represent the two geographic
locations. The letters represent different pre-defined
habitat types (A, B and C). If the samples are arranged
randomly, there is no effect of either current ecology or
past history at the taxonomic and spatial resolution sampled, indicating that all the samples were taken within
one microbial habitat and one province. Alternatively,
biotic composition might cluster into multiple microbial
habitats, geographic locations indicating multiple provinces, or both. Various statistical methods can test for
significant patterns, such as a two-factor clustering test
or canonical analysis35.
If the samples cluster by habitat, it can be concluded
that the assemblages are influenced by the contemporary environment. But what does it mean if geographic
separation influences biotic composition? The key to
interpreting this result is to be able to determine whether
isolation by distance (or a geographic barrier) influences composition even after controlling for presentday environmental factors. Of course, the pre-defined
habitats might not capture all possible contemporary
environmental variation among sample sites, but with
good replication (that is, sampling the same habitat
types in many different geographic locations), one can
be increasingly certain of a distance effect. Such a distance
effect is strong evidence of biogeographic provincialism,
in which differences in biotic composition are due to
past events rather than present-day attributes of the
environment. Although there is no direct effect of distance per se, distance is related to the likelihood that past
divergence of biotic assemblages, whether due to genetic
drift or adaptation to past environments, is maintained
by genetic isolation.
Often, researchers record continuous variables to
describe the environment and spatial arrangement of
their sampling sites rather than using discrete categories. The above analysis can be modified to incorporate
continuous measures by deriving two additional matrices: an environmental-similarity matrix (including both
biotic and abiotic variables) and a geographic-distance
matrix (BOX 1). The correlation, or lack thereof, between
the three summary matrices in BOX 1 can then be used
to distinguish the four hypotheses (FIG. 1b).
NATURE REVIEWS | MICROBIOLOGY
VOLUME 4 | FEBRUARY 2006 | 105
© 2006 Nature Publishing Group
REVIEWS
a
One microbial
province
One microbial
habitat
Effects of environment and history
Whereas many studies examine whether microbial
assemblages vary over space, we know of only 10 published microbial studies that can be applied to the above
framework to assess the roles of historical contingencies
A
C
C
A
B
B
B
C
C
A
A
B
B
A
B
B
B
One microbial
province
Environmental similarity ( )
Geographic distance ( )
Environmental similarity ( )
Geographic distance ( )
Community similarity
B
A
A
C
A
B
Community similarity
A
B
C
C
A
A
B
C
B
C
Community similarity
C
A
C
B
A
Community similarity
A
B
A
C
Multiple microbial
habitats
One microbial
habitat
Multiple microbial
habitats
C
B
A
b
Multiple microbial
provinces
C
C
C
Multiple microbial
provinces
Environmental similarity ( )
Geographic distance ( )
Environmental similarity ( )
Geographic distance ( )
Figure 1 | Assessing the contributions of
environmental and historical effects on microbial
biogeography. Four alternative hypotheses about
environmental and historical influences on
communities and the general results that would
support them, using (a) samples (shown in circles) from
discrete predefined habitat types (A, B and C) and
locations (green versus white) or (b) samples from
continuous habitat variables and geographic distances.
The axes in (a) are dimensionless; samples that contain
similar assemblages are mapped closer to one another
relative to pairs of samples with different assemblages.
In b, lack of a correlation between environmental
similarity or geographic distance and biotic similarity
(BOX 1) indicates no biogeographic patterning.
Alternatively, to the extent that environmental and
historical factors have influenced the assemblages
sampled, biotic similarity should be correlated with
environmental similarity and geographic distance,
respectively. Standard correlation tests are not
appropriate to distinguish between these hypotheses
because of non-independence; therefore,
randomization tests such as a bootstrapped regression
analysis25 or Mantel tests87,88 are required. Further tests,
such as a partial Mantel test, can disentangle the effects
of geographic distance versus environment on
assemblage composition89.
and contemporary environmental factors (TABLE 2). The
number of studies available is small because few microbial biogeography studies report the geographic distance
between their samples or directly test for a distance effect
relative to a contemporary environmental effect.
Despite this low number, five studies found significant distance effects, indicating at least some degree of
provincialism (TABLE 2). The relative influence of historical versus environmental factors seems to be related to
the scale of sampling. In the two intercontinental studies, Synechococcus36 and Sulfolobus37 assemblages in hot
springs could be significantly differentiated by distance
but did not correlate with the many environmental
variables measured. This result indicates that, on the
order of tens of thousands of kilometers, the legacy of
historical separation can overwhelm any effect of environmental factors. These results should not be taken
to mean that the contemporary environment has no
effect on biotic composition, but rather that its influence is relatively small compared to that of distance.
Indeed, Papke et al.36 noted that, although there was
no strong correlation between Synechococcus genotype
and chemical characteristics of hot springs at an intercontinental scale, some genetic differences among hot
springs within a continent seemed to be due to spring
chemistry.
By contrast, environmental effects have been repeatedly shown to significantly influence biotic composition
at small spatial scales for which distance effects seem to
be negligible. The two studies that sampled sites separated
by only a few kilometers found significant environmental effects but none of distance (TABLE 2). At intermediate
scales (10–3000 km), three of the five studies found a
significant distance effect. Environmental conditions
also seemed to influence composition at this spatial scale,
with one exception38. Therefore, it is at this intermediate
spatial scale that the influence of both historical contingencies and contemporary ecological factors on microbial
biogeography is most likely to be detected.
These general trends were apparent even though the
studies included a broad range of taxa (bacteria, archaea
and fungi) and the resolution of the taxonomic units varied enormously. The OTU (operational taxonomic unit)
definitions varied from ARISA (automated ribosomal
intergenic spacer analysis) profiles of all bacteria to multilocus sequence typing of cultured isolates within one
genus. In general, patterns present at finer taxonomic
resolutions might not be reflected at broader resolutions.
For instance, one might not see compositional differences among sites at the level of 16S rDNA, even though
there are clear differences in ITS (intergenic transcribed
space) sequences20. The same taxonomic-scale dependence applies to the distribution of macroorganisms; for
instance, many more plant genera are restricted to a
particular continent than plant families.
What processes shape microbial biogeography?
The studies reviewed above indicate that microbial
assemblages can exhibit both environmental segregation and biogeographic provincialism (TABLE 2), but
what processes generate these patterns? The definitive
106 | FEBRUARY 2006 | VOLUME 4
www.nature.com/reviews/micro
© 2006 Nature Publishing Group
REVIEWS
Table 2 | Studies of the effects of distance (dist.) and environment (env.) on microbial composition
Allometry
The relationship between
organismal attributes and body
size of the form Y = Y 0 Mb , in
which Y is a variable such as
metabolic rate, lifespan or
population density, Y 0 is a
normalization constant (the
y-intercept on a logarithmic
graph), M is body mass (or
other measure of body size)
and b is the scaling exponent
(the slope on the graph).
Ecological drift
The influence of random
demographic variability (such
as birth, death and migration
rates) on biotic composition.
Propagule
The smallest unit of dispersal
that is necessary to colonize a
new population.
Organisms
Approximate
scale (km)
Habitat
OTU
Synechococcus
20,000
Hot springs
16S/ITS sequence
Sulfolobus
12,000
Hot springs
MLS of isolates
Yes*
No*
37
Bacteria
3,000
Coral
16S sequence
No
Yes*
108
Bacteria
500
Lakes
ARISA
Yes*
Yes*
109
3-CBD bacteria
500
Soil
ARDRA
No
Yes*
87
Log (lifetime dispersal capability)
36
Soil
ARISA
Yes*
Yes*
26
Aquatic
ARISA
No
Yes
110
Bacteria
10
Lakes
DGGE of 16S
Yes*
No*
38
Bacteria
0.3
Marsh sediment
16S sequence
No*
Yes*
27
Bacteria
0.1
Soil
TRFLP
No
Yes*
33
The studies are ordered by the geographical scale over which the samples were taken, reported as the approximate furthest
distance between sampling points. *The effect was tested for statistical significance. 3-CBD, 3-chlorobenzoate-degrading;
ARDRA, amplified ribosomal DNA restriction analysis; ARISA, automated ribosomal intergenic spacer analysis; DGGE, denaturing
gradient gel electrophoresis; ITS, intergenic transcribed space; MLS, multilocus sequencing; OTU, operational taxonomic unit used
in the study; TRFLP, terminal RFLP.
Global
In
No
100
Micrometers
zed
e-si s
diat ganism
e
m
or
ter
env.
Yes
100
b Passive dispersal
ms
anis
dist.
Bacteria
a Active dispersal
org
Ref.
Ascomycetes
difference between all micro- and macroorganisms is
their size. As a first hypothesis, we suggest that the same
processes that influence macroorganism biogeography
also apply to microbial life but that their rates scale with
body size or, for single-celled organisms, cell size. Many
attributes, from metabolic rate to maximum lifespan,
vary predictably with an organism’s size. Often, this variation yields linear relationships on a logarithmic plot and
therefore can be described by a so-called allometric equation. This idea of allometry39 serves as a useful structure
for the discussion of biogeographic processes below.
ro
Mic
Effect of
ants
e pl ls
Larg anima
and
Log (mass)
ms
zed
e-si s
diat ganism
e
m
or
nter
anis
org
ro
Mic
I
ants
e pl ls
Larg anima
and
Log (mass)
Figure 2 | Hypothetical relationship between body mass (at an organism’s largest
life stage) and lifetime dispersal capability. The relationship varies depending on
whether the organism disperses actively or passively (by its own propulsion). The range of
active (a) dispersal is a subset of the range of passive (b) dispersal. It is convenient to think
of the log(mass) axis as representing three qualitative groups: first, microorganisms,
which span about 8 orders of magnitude from bacteria to eukaryotic algae and protozoa
(10–13–10–5 g); second, large plants and animals, which span about 8 orders of magnitude
from herbs and small vertebrates to whales and trees (101–109 g); and third, intermediatesized organisms, which span the intervening 6 or so orders of magnitude and include the
small metazoans, such as nematodes, annelids and arthropods.
For plant and animal biogeography, the most relevant
rates are those processes by which a taxon expands or
contracts its area of distribution1. We discuss three such
processes: colonization, speciation (generally, diversification) and extinction.
Colonization. One of the main arguments behind the
‘everything is everywhere’ hypothesis of microbial biogeography is that the dispersal and subsequent colonization of microorganisms into new locations is so great
that it prevents spatial differentiation. High dispersal
rates decrease assemblage differentiation by increasing
gene flow, whether by means of sexual reproduction or
horizontal gene transfer, overwhelming any tendency
towards genetic differentiation due to mutation, selection or genetic drift1; and by mixing of individuals, overwhelming spatial differences in taxon abundance due to
ecological interactions or to ecological drift40,41.
From the perspective of allometry, the question is:
does an organism’s size influence its dispersal ability?
Of particular interest are long-distance dispersal events
that transport a propagule across barriers of inhospitable
habitats. FIGURE 2 shows the hypothesized relationship
between body mass and dispersal capability, which is the
maximal distance traveled by an individual in its lifetime
(or between cell divisions). This relationship depends on
whether an organism disperses primarily by active propulsion, such as propelling itself micrometers with its flagella,
or by passive transport, such as being carried thousands of
kilometers by ocean currents or migrating birds.
Several points are immediately apparent from FIG. 2.
First, dispersal capacity is one of the least likely attributes
to be constrained by the size of an organism and is not
well characterized by an allometric equation. We can,
however, hypothesize about the potential constraints of
this relationship. Second, there are severe constraints on
microbial dispersal by active propulsion (FIG. 2a). Large
organisms range from having little or no active dispersal (trees, giant clams and corals) to dispersing over
NATURE REVIEWS | MICROBIOLOGY
VOLUME 4 | FEBRUARY 2006 | 107
© 2006 Nature Publishing Group
Probability of dispersal
Local
Global
Distance from origin (meters)
Figure 3 | Hypothetical dispersal distribution of a
typical passively dispersed macroorganism.
Population density influences the probability that an
individual from that population will disperse over very
long distances (solid line). For taxa with relatively low
densities (dashed line), dispersal might be effectively
restricted, even though long-distance movement is
theoretically possible. Based on REF. 41.
Log (density)
Log (density)
Phytoplankton
2.20
0.14
0.20
4.10
Log (body mass)
Birds
Low
thousands of kilometers (whales, birds and butterflies).
By contrast, microorganisms have little capacity to cross
significant geographic barriers under their own propulsion. Of course, over many generations, a bacterial taxon
could eventually spread great distances by active propulsion. By that time, however, genetic divergence from the
source population would likely occur, thereby generating
biogeographic structure rather than eliminating it. Third,
there seem to be no size constraints on passive dispersal
(FIG. 2b). Some large organisms, such as elephants and
rhinoceroses, have negligible passive dispersal, whereas
others, such as tree ferns, trees and giant clams, disperse
thousands of kilometers by air or water as spores, seeds
and larvae. We extrapolate that passive dispersal by microorganisms is equally broad — whereas some microbial
taxa might disperse globally, others will only disperse over
very short distances, creating non-random distributions
of microbial assemblages.
What factors might limit the passive dispersal of a
microorganism (and fill in the bottom-left-hand corner
of FIG. 2b)? Certainly, habitat will have a role. Cells in
subsurface soils and sediment will not disperse as far as
those in water and surface soils. Furthermore, the propagule must survive the conditions encountered during
dispersal to a new suitable location. The ‘everything is
everywhere’ dictum implicitly assumes that all microorganisms are highly tolerant to stress; however, not all
microorganisms produce spores and cysts, and those
that do vary greatly in their hardiness. Last, to grow
and establish a population, the propagule must be able
to outcompete local populations that might be better
adapted to the specific conditions.
A taxon’s colonization rate depends on its population
density as well as its dispersal ability. This relationship
is best illustrated by a dispersal frequency distribution (FIG. 3). For passively dispersed macroorganisms,
High
REVIEWS
Microorganisms
Intermediatesized organisms
Large plants
and animals
Log (mass)
Figure 4 | Hypothesized constraints on a taxon’s
population density in a given body-size class. The
thick green line on the diagonal is a known physiological
constraint. The gradient in shading from the diagonal to
the bottom-left corner represents the idea that fewer taxa
are thought to fall in the bottom left of the figure; however,
we hypothesize that some taxa do fall in this region. The
inset plots log (body mass, g) of North American birds
versus log (population density, individuals per route). The
data set falls within a well-defined quadrilateral with a
constant minimum density and a maximum density for
birds of an intermediate size. Individual data points are not
shown. The outline of these data is also sketched on the
constraint figure. The approximate range of marine
phytoplankton data from Li48 is also sketched (assuming
that cell volume is proportional to body mass). The X-axis
categories are defined in FIG. 2. Inset adapted with
permission from REF. 43 © (1987) University of Chicago.
most propagules move only very short distances, but
a small proportion can disperse over vast distances
(for example, in seed dispersal)42. As similar processes
contribute to their dispersal, we hypothesize that the
shape of the frequency distribution of microbial dispersal distances is similar to that observed with passively
dispersed macroorganisms. In general, large numbers
of potential propagules increase the chance that at
least one will travel a long distance and establish a new
persistent population, whereas low densities effectively
shorten the tail of the dispersal distribution (FIG. 3).
Given that microorganisms have finite population sizes,
the low-probability, long-distance dispersal events that
are expected to occur eventually by chance might occur
rarely or not at all.
How do the population densities of microbial taxa
compare to those of macroorganisms? FIGURE 4 illustrates the relationship observed in macroorganisms and
extends these observations to microorganisms. There are
three important features to note here. First, on average,
there seems to be a negative relationship between a taxon’s size and its population density. In many studies43–46,
smaller-sized organisms have, on average, larger population densities than bigger organisms (although the maximum population density is often not the smallest size
category)47. If the same relationship holds at the scale of
microbial sizes, then, on average, microorganisms will
have larger population densities than macroorganisms.
Indeed, Li48 found a negative relationship between the
108 | FEBRUARY 2006 | VOLUME 4
www.nature.com/reviews/micro
© 2006 Nature Publishing Group
REVIEWS
Geographic range
The area encompassing the
extent of a taxon’s distribution.
Global
The balance of biogeographic processes. Ultimately, it is
a balance between origination and extinction processes
that determines global taxon diversity and shapes biogeography. There is evidence that the balance of global species
diversity and, therefore, the underlying processes scale
with organism size (FIG. 6; for examples, see REFS 63,64).
The exact relationship is still unclear, however, and the
number of species might decrease with decreasing size at
the lower limit of the size distribution of macroorganisms.
This decrease might be an artefact owing to undersampling and taxonomic lumping of small organisms or,
alternatively, it could be the true relationship.
These two alternatives predict large differences in
total microbial diversity. If we assume that the apparent
decrease in richness of the smaller macroorganisms is an
artefact and that the pattern for larger macroorganisms
reflects the true relationship, then microbial diversity
within a size class is predicted to increase allometrically
with decreasing size. As such, microbial diversity would
be much higher than the diversity of macroorganisms
(as suggested by Dykhuizen65). By contrast, if richness
peaks at some intermediate body size (as suggested in
FIG. 6), then total microbial diversity might be lower than
total macroorganism diversity (as suggested by Finlay9).
Of course, a complication in comparing the biogeography of micro- and macroorganisms is the problem of
comparing equivalent taxonomic units. Not only would
sufficient sampling be needed to assess the global diversity of microorganisms within different size classes, but
their diversity would need to be measured using taxon
definitions comparable to those of plants and animals.
Birds
Log
(geographicrange size)
Log (geographic-range size)
Diversification and extinction. Diversification owing to
mutation, genetic drift and differential selective pressures
will generate biogeographic patterns, unless it is counterbalanced by the forces of dispersal and homogenizing
selection. Evidence is mixed as to whether speciation
rates are related to body size in macroorganisms50,51. If
laboratory studies are any indication52,53, the potential for
rapid diversification seems greater in microorganisms
than macroorganisms. Microbial species typically have
higher densities (FIG. 4) and shorter generation times
than macroorganisms, allowing them to undergo rapid
genetic divergence.
Extinction influences biogeographic patterns, usually by eliminating endemic forms. Species with small
geographic-range sizes have a higher probability of extinction54,55; large geographic ranges provide insurance
against extinction owing to local disturbances. In macroorganisms, there is a weak but significant positive correlation between body size and range size (FIG. 5), and the
relationship is better characterized as a constraint space.
The upper limit on range size extends to the entire globe
and is independent of body mass, as examples of global
taxa are found among bacteria49 as well as plants and animals1,56. By contrast, the lower limit of range size probably
depends on organism size. There are few large organisms
that occupy small ranges (FIG. 5), probably because the
lower population density of larger organisms (FIG. 4) results
in an increased likelihood of extinction43,57. Indeed, many
large-bodied mammals and birds that occupy limited
ranges are recently extinct or are currently endangered58.
How might range sizes vary within the constraints of
FIG. 5, and therefore influence microbial extinction rates?
The fact that there are some ‘cosmopolitan’ microorganisms (for examples, see REFS 59-61) should not be taken to
imply that most taxa are so widely distributed. The modal
range size of macroorganisms within a taxonomic group
tends to be intermediate — that is, most species are neither
extremely narrowly nor globally distributed62. If the pattern in macroorganisms extends to microorganisms, we
expect that range sizes within microbial-cell size classes
vary greatly. Variation in range sizes indicates variation in
extinction rates; those microorganisms that have relatively
restricted distributions should have higher extinction rates
than those that are more global in range.
Micrometers
cell volume of marine phytoplankton and their population densities (FIG. 4). Second, there is no reason to think
that small organisms that have low population densities do not exist (in other words, there are taxa in the
bottom-left-hand corner of FIG. 4). Indeed, this must be
true if, for no other reason, the population density of a
microbial taxon must be low following a diversification
event or just prior to extinction. Last, there are theoretical and empirical reasons to expect that the upper limit
to population density for a given size class increases as
size decreases (the thick line in FIG. 4). Smaller organisms
need fewer resources per individual and can therefore
have higher population densities in an area.
In conclusion, the combination of dispersal ability and
population density of a taxon determines its rate of colonizing new and distant habitats. Whereas some microbial taxa could possess the combination of traits that
allows them to colonize at a global scale (spore-forming
Bacillus49, for example), others might have short dispersal distances and restricted geographic distributions
(hot-spring Sulfolobus37, for example).
1.40
–1.60
0.20
4.10
Log (body mass)
Microorganisms
Intermediatesized organisms
Large plants
and animals
Log (mass)
Figure 5 | Hypothesized constraints on an organism’s
geographic-range size for a given body mass. The inset
graph is log (body size, g) versus log (geographic-range size,
106 km2) for terrestrial bird species of North America.
Individual data points are not shown. The combined data
set forms an approximate triangle. The outline of these data
is also sketched on the constraint figure. The X-axis
categories are defined in FIG. 2. Inset adapted with
permission from REF. 43 © (1987) University of Chicago.
NATURE REVIEWS | MICROBIOLOGY
VOLUME 4 | FEBRUARY 2006 | 109
© 2006 Nature Publishing Group
REVIEWS
Determining the nature of this relationship would
provide insight into the nature of the processes underlying biogeographic patterns. If the relationship between
global diversity and body size proves to be hump-shaped,
this indicates that diversification and extinction do not
both scale allometrically with body size, and that a
body-size threshold might exist across which the balance
between diversification and extinction is determined by
fundamentally different factors.
Number of species
42
30
18
6
–8 –6 –4 –2 0
Log (body mass, g)
Low
Log (species number)
High
Discussion
A large body of research supports the idea that freeliving microorganisms exhibit biogeographic patterns.
Current evidence confirms that, as Baas-Becking proposed, the environment selects and is, in part, responsible for spatial variation in microbial diversity. However,
recent studies dispute the idea that ‘everything is everywhere’. Instead, the legacies of historical events have
left lasting signatures on the distributions of microbial
assemblages, even at distances as small as 500 km.
These results indicate that there are some aspects
of biogeography that might be common to all of life.
However, there are other aspects of biogeography that
might be unique to microorganisms. For example, we
conclude that the rates of the processes underlying
biogeography probably vary more widely for microorganisms of a given size than for macroorganisms of
a given size. Except for the case of active dispersal, we
hypothesize that body size does not constrain a microorganism’s dispersal rate, population density and range
size, whereas it does somewhat constrain those of a
larger organisms (FIGS 2,4,5). Therefore, a question for
future research is: what are the traits that lead to the wide
variety of colonization, diversification and extinction
rates in microorganisms (and which at the same time
Microorganisms
Intermediate-sized
organisms
Large plants
and animals
Log (mass)
Figure 6 | Hypothesized relationships between number
of species and body mass. For larger macroorganisms
(approximately larger than insects), it is clear that the
number of species increases as body mass decreases. For
smaller macroorganisms and microorganisms, the number
of species might continue to increase (dashed line) or begin
to decrease (dotted line) as body mass decreases. The inset
plots the number of invertebrate species by log(mass) on
Marion Island64. The X-axis categories are defined in FIG. 2.
Inset reproduced with permission from REF. 64 © (2001)
National Academy of Sciences, USA.
are relatively more constrained for macroorganisms)?
Our discussion was limited by treating biogeographic
processes separately; further work is needed to assess the
relative importance of these processes for different types
of microorganisms.
An unavoidable problem in comparing the patterns of
micro- and macroorganisms is how to compare equivalent
taxonomic units. In theory, the same taxonomic resolution should be used, whether that unit is a species or a
sequence-similarity cut-off. In practice, most macroorganism taxa, species or otherwise are defined by morphological characteristics, which are not tightly correlated with
genetic differentiation. For instance, the degree of genetic
variation of fish, bird and mammal species within genera
spans two orders of magnitude among these genera66.
Furthermore, the level of taxonomic resolution in
microbial-diversity studies is generally much coarser than
that adopted with macroorganisms. For example, human
and chimpanzee genomes exhibit 98.6% DNA–DNA
hybridization between them, yet bacterial systematists have traditionally defined a species as strains with
genomes that exhibit over 70% DNA–DNA hybridization8.
Similarly, microbial eukaryotes are often differentiated by
morphological characteristics, but their small size means
that few characteristics can be distinguished. Therefore, a
single Latin binomial for a microbial eukaryote can often
refer to a complex of cryptic species67. Consequently,
declaring that these ‘species’ are cosmopolitan, for example7,9, might be approximately equivalent to saying that a
genus or family of birds is cosmopolitan.
Just as the long-standing debate about species
definitions remains unresolved for microorganisms68
and macroorganisms69, so will the broader question of
whether, and how, to compare them. We are optimistic,
however. Despite the difficulty of defining microbial
taxonomic units, biogeographic patterns seem robust
enough to be detectable across various taxa (TABLE 1).
It is also possible to avoid comparing taxonomic units
by asking instead whether there is a level of taxonomic
resolution at which microbial biogeographic patterns
approach those of macroorganisms70.
The four hypotheses illustrated in FIG. 1 provide a useful framework to further explore microbial biogeography,
beyond merely documenting the existence of patterns.
Many microbial data sets have already been collected that
could be interpreted within this framework. Many, however, cannot, primarily because such studies have failed
to take samples separated by a range of known distances,
to sample across various habitats, to gather contemporaneous environmental data or all of the above. In general,
designing and organizing data sets to correlate microbial
genomic and metagenomic data with ecological and environmental information can provide biological insights71.
Specifically, we recommend that new microbial biogeography studies should systematically sample and record
data from various distances, habitats and environmental
conditions, to better distinguish between contemporary
and historical factors. If they do not, the field of microbial
biogeography will probably become mired in phenomenological description, instead of tackling the mechanisms
that generate the patterns.
110 | FEBRUARY 2006 | VOLUME 4
www.nature.com/reviews/micro
© 2006 Nature Publishing Group
REVIEWS
The motivation for understanding microbial biogeography extends beyond drawing and interpreting
a map of microbial diversity. As with macroorganisms (for example, see REFS 72–75), a growing body
of evidence indicates that microbial composition also
affects ecosystem processes, including CO2 respiration
and decomposition76,77, autotrophic and heterotrophic
production78,79 and nitrogen cycling80,81. Therefore, even
under similar environmental conditions, microbial communities from different provinces might function differently. A better understanding of microbial biogeography
is essential to predict such effects. It is also crucial in the
search for novel pharmaceuticals and other compounds
of industrial importance82.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
Brown, J. H. & Lomolino, M. V. Biogeography
(Sinauer, Sunderland, 1998).
A definitive textbook on the biogeography of
macroorganisms; has just been updated in a 2005
edition.
Ward, D. M., Weller, R. & Bateson, M. M. 16S
ribosomal-RNA sequences reveal numerous
uncultured microorganisms in a natural community.
Nature 345, 63–65 (1990).
Øvreås, L. Population and community level
approaches for analysing microbial diversity in natural
environments. Ecol. Lett. 3, 236–251 (2000).
Floyd, M. M., Tang, J., Kane, M. & Emerson, D.
Captured diversity in a culture collection: case study of
the geographic and habitat distributions of
environmental isolates held at the American type
culture collection. Appl. Environ. Microbiol. 71,
2813–2823 (2005).
Schloss, P. D. & Handelsman, J. Status of the
microbial census. Microbiol. Mol. Biol. Rev. 68,
686–691 (2004).
Venter, J. C. et al. Environmental genome shotgun
sequencing of the Sargasso Sea. Science 304, 66–74
(2004).
Fenchel, T., Esteban, G. F. & Finlay, B. J. Local versus
global diversity of microorganisms: cryptic diversity of
ciliated protozoa. Oikos 80, 220–225 (1997).
Staley, J. T. Biodiversity: are microbial species
threatened? Curr. Opin. Biotechnol. 8, 340–345
(1997).
Finlay, B. J. Global dispersal of free-living microbial
eukaryote species. Science 296, 1061–1063 (2002).
A summary of the arguments for why microbial
eukaryotes might not be restricted by geographic
barriers.
Hedlund, B. P. & Staley, J. T. Microbial endemism and
biogeography. In Microbial Diversity and Bioprospecting
(ed. Bull, A. T.) (ASM, Washington DC, 2003).
Anagnostakis, S. Chestnut blight: the classical
problem of an introduced pathogen. Mycologia 79,
23–27 (1987).
Falush, D. et al. Traces of human migrations in
Helicobacter pylori populations. Science 299,
1582–1585 (2003).
Breitbart, M. & Rohwer, F. Here a virus, there a virus,
everywhere the same virus? Trends Microbiol. 13,
278–284 (2005).
de Candolle, A. P. Essai Elementaire de Geographie
Botanique (F. G. Levrault, Paris, 1820).
Beijerinck, M. W. De infusies en de ontdekking der
backteriën. In Jaarboek van de Koninklijke Akademie
van Wetenschappen (Müller, Amsterdam, 1913).
Baas-Becking, L. G. M. Geobiologie of Inleiding Tot de
Milieukunde (Van Stockkum & Zoon, The Hague, 1934).
Zhang, N. & Blackwell, M. Population structure of
dogwood anthracnose fungus. Phytopathology 92,
1276–1283 (2002).
Bala, A., Murphy, P. & Giller, K. E. Distribution and
diversity of rhizobia nodulating agroforestry
legumes in soils from three continents in the tropics.
Mol. Ecol. 12, 917–929 (2003).
Papke, R. T. & Ward, D M. The importance of
physical isolation to microbial diversification. FEMS
Microbiol. Ecol. 48, 293–303 (2004).
A recent review discussing how physical isolation
might affect prokaryote evolution.
If our initial biogeographic hypotheses based on extrapolating from macroorganisms to microorganisms are
eventually rejected, this outcome would support the view
that “biodiversity at the microbial level is fundamentally
different from that of macroscopic animals and plants”83.
Discovering exactly which attributes and processes contribute to these fundamental differences would greatly
further our understanding of all living things. If, however,
biogeographic and allometric patterns and processes are
found to be fundamentally similar in all organisms, this
will provide yet another example of the unity of life — the
extent to which all living things not only use similar molecules for structure and function, but also follow similar
ecological, evolutionary and biogeographic principles.
20. Cho, J. C. & Tiedje, J. M. Biogeography and degree
of endemicity of fluorescent Pseudomonas strains in
soil. Appl. Environ. Microbiol. 66, 5448–5456
(2000).
One of the first examples of relating the genetic
similarity of a free-living bacterial assemblage
with geographic distance, using fluorescent
Pseudomonas isolates.
21. Oda, Y., Star, B., Huisman, L. A., Gottschal, J. C.
& Forney, L. J. Biogeography of the purple nonsulfur
bacterium Rhodopseudomonas palustris. Appl.
Environ. Microbiol. 69, 5186–5191 (2003).
22. Crump, B. C., Hopkinson, C. S., Sogin, M. L. &
Hobbie, J. E. Microbial biogeography along an
estuarine salinity gradient: combined influences of
bacterial growth and residence time. Appl. Environ.
Microbiol. 70, 1494–1505 (2004).
23. Øvreås, L., Forney, L., Daae, R. L. & Torsvik, V.
Distribution of bacterioplankton in meromictic Lake
Sælenvannet, as determined by denaturing gradient
gel electrophoresis of PCR-amplified gene fragments
coding for 16S rRNA. Appl. Environ. Microbiol. 63,
3367–3373 (1997).
24. Schwalbach, M. S., Hewson, I. & Fuhrman, J. A.
Viral effects on bacterial community composition in
marine plankton microcosms. Aquat. Microb. Ecol.
34, 117–127 (2004).
25. Rosenzweig, M. L. Species Diversity in Space and
Time. (Cambridge University Press, Cambridge, 1995).
26. Green, J. L. et al. Spatial scaling of microbial
eukaryote diversity. Nature 432, 747–750 (2004).
Along with reference 27, this study uses
distance–decay curves to demonstrate nonrandom bacterial-assemblage distributions; the
studies then test whether the non-random
distributions are due to isolation by distance
and/or local environmental conditions.
27. Horner-Devine, M. C., Lage, M., Hughes, J. B. &
Bohannan, B. J. M. A taxa–area relationship for
bacteria. Nature 432, 750–753 (2004).
28. Noguez, A. M. et al. Microbial macroecology: highly
structured prokaryotic soil assemblages in a tropical
deciduous forest. Glob. Ecol. Biogeogr. 14,
241–248 (2005).
29. Bell, T. et al. Larger islands house more bacterial
taxa. Science 308, 1884 (2005).
30. Smith, V. H. et al. Phytoplankton species richness
scales consistently from laboratory microcosms to
the world’s oceans. Proc. Natl Acad. Sci. USA 102,
4393–4396 (2005).
31. Cam, E. et al. Disentangling sampling and ecological
explanations underlying species–area relationships.
Ecology 84, 1118–1130 (2002).
32. Colwell, R. K., Mao, C. X. & Chang, J. Interpolating,
extrapolating, and comparing incidence-based
species accumulation curves. Ecology 85,
2717–2727 (2004).
33. Kuske, C. R. et al. Comparison of soil bacterial
communities in rhizospheres of three plant species
and the interspaces in an arid grassland. Appl.
Environ. Microbiol. 68, 1854–1863 (2002).
34. Yannarell, A. C. & Triplett, E. W. Within- and
between-lake variability in the composition of
bacterioplankton communities: investigations using
multiple spatial scales. Appl. Environ. Microbiol. 70,
214–223 (2004).
NATURE REVIEWS | MICROBIOLOGY
35. Legendre, P. & Legendre, L. Numerical Ecology 2nd
edn (Elsevier, Amsterdam, 1998).
36. Papke, R. T., Ramsing, N. B., Bateson, M. M. &
Ward, D. M. Geographical isolation in hot spring
cyanobacteria. Environ. Microbiol. 5, 650–659 (2003).
37. Whitaker, R. J., Grogan, D. W. & Taylor, J. W.
Geographic barriers isolate endemic populations of
hyperthermophilic Archaea. Science 301, 976–978
(2003).
Already a classic microbial biogeography study, it
considers both the effects of spatial isolation and
contemporary environmental parameters on
hotspring Sulfolobus assemblages within and
across continents.
38. Reche, I., Pulido-Villena, E., Morales-Baquero, R.
& Casamayor, E. O. Does ecosystem size
determine aquatic bacterial richness? Ecology
86, 1715–1722 (2005).
39. Brown, J. H., West, G. B. & Enquist, B. J. Scaling in
biology: patterns and processes, causes and
consequences. In Scaling in Biology (eds Brown, J.
H. & West, G. B.) 1–24 (Oxford University Press,
Oxford, 2000).
A comprehensive introduction to allometric
patterns in biology.
40. Bell, G. The distribution of abundance in neutral
communities. Am. Nat. 155, 606–617 (2000).
41. Hubbell, S. P. The Unified Neutral Theory of
Biodiversity and Biogeography (Princeton University
Press, Princeton, 2001).
42. Clark, J. S., Silman, M., Kern, R., Macklin, E. &
HilleRisLambers, J. Seed dispersal near and far:
patterns across temperate and tropical forests.
Ecology 80, 1475–1494 (1999).
43. Brown, J. H. & Maurer, B. A. Evolution of species
assemblages: effects of energetic constraints and
species dynamics on the diversification of the North
American avifauna. Am. Nat. 130, 1–17 (1987).
44. Morse, D. R., Stork, N. E. & Lawton, J. H. Species
number, species abundance and body length
relationships of arboreal beetles in Bornean lowland
rain forest trees. Ecol. Entomol. 13, 25–37 (1988).
45. Lawton, J. H. Species richness and population
dynamics of animal assemblages. Patterns in body
size: abundance space. Philos. Trans. R. Soc. Lond. B
Biol. Sci. 330, 283–291 (1990).
46. Siemann, E., Tilman, D. & Haarstad, J. Insect species
diversity, abundance and body size relationships.
Nature 380, 704–706 (1996).
47. Brown, J. H. Macroecology (Chicago University Press,
Chicago, 1995).
48. Li, W. K. W. Macroecological patterns of
phytoplankton in the northwestern North Atlantic
Ocean. Nature 419, 154–157 (2002).
49. Roberts, M. S. & Cohan, F. M. Recombination and
migration rates in natural populations of Bacillus
subtilis and Bacillus mojavensis. Evolution 49,
1081–1094 (1995).
50. Stanley, S. M. Macroevolution: Pattern and Process
(W. H. Freeman, San Francisco, 1979).
51. Paradis, E. Statistical analysis of diversification with
species traits. Evolution 59, 1–12 (2005).
52. Lenski, R. E. & Travisano, M. Dynamics of adaptation
and diversification: a 10,000-generation experiment
with bacterial-populations. Proc. Natl Acad. Sci. USA
91, 6808–6814 (1994).
VOLUME 4 | FEBRUARY 2006 | 111
© 2006 Nature Publishing Group
REVIEWS
53. Rainey, P. B. & Travisano, M. Adaptive radiation in a
heterogeneous environment. Nature 394, 69–72
(1998).
54. Manne, L. L., Brooks, T. M. & Pimm, S. L. Relative risk
of extinction of passerine birds on continents and
islands. Nature 399, 258–261 (1999).
55. Purvis, A., Gittleman, J. L., Cowlishaw, G. & Mace, G.
M. Predicting extinction risk in declining species. Proc.
R. Soc. Lond. B Biol. Sci. 267, 1947–1952 (2000).
56. Lawton, J. H. Range, Population abundance and
conservation. Trends Ecol. Evol. 8, 409–413 (1993).
57. Silva, M. & Downing, J. A. Allometric scaling of
minimal mammal densities. Conserv. Biol. 8, 732–
743 (1994).
58. Meretsky, V. J., Snyder, N. F. R., Beissinger, S. R.,
Clendenen, D. A. & Wiley, J. W. Demography of the
California condor: implications for reestablishment.
Conserv. Biol. 14, 957–967 (2000).
59. Finlay, B. J. & Clarke, K. J. Apparent global ubiquity of
species in the protist genus Paraphysomonas. Protist
150, 419–430 (1999).
60. Massana, R., DeLong, E. F.& Pedrós-Alió, C. A few
cosmopolitan phylotypes dominate planktonic archaeal
assemblages in widely different oceanic provinces.
Appl. Environ. Microbiol. 66, 1777–1787 (2000).
61. Lindström, E. S. & Leskinen, E. Do neighboring lakes
share common taxa of bacterioplankton? Comparison
of 16S rDNA fingerprints and sequences from three
geographic regions. Microb. Ecol. 44, 1–9 (2002).
62. Brown, J. H., Stevens, G. C. & Kaufman, D. M. The
geographic range: size, shape, boundaries, and
internal structure. Annu. Rev. Ecol. Syst. 27, 597–
623 (1996).
63. May, R. M. How many species are there on Earth?
Science 241, 1441–1449 (1988).
64. Gaston, K. J., Chown, S. L. & Mercer, R. D. The animal
species-body size distribution of Marion Island. Proc.
Natl Acad. Sci. USA 98, 14493–14496 (2001).
65. Dykhuizen, D. E. Santa Rosalia revisited: why are there
so many species of bacteria? Antonie Van
Leeuwenhoek 73, 25–33 (1998).
66. Avise, J. C. & Aquadro, C. F. A comparative summary
of genetic distances in the vertebrates. Evol. Biol. 15,
151–185 (1982).
67. de Vargas, C., Norris, R., Zaninetti, L., Gibb, S. W. &
Pawlowski, J. Molecular evidence of cryptic speciation
in planktonic foraminifers and their relation to oceanic
provinces. Proc. Natl Acad. Sci. USA 96, 2864–2868
(1999).
68. Cohan, F. M. What are bacterial species? Annu. Rev.
Microbiol. 56, 457–487 (2002).
69. Coyne, J. A., Orr, H. A. & Futuyma, D. J. Do we need a
new species concept? Syst. Zool. 37, 190–200
(1988).
70. Green, J. L. & Bohannan, B. J. M. Spatial scaling of
microbial biodiversity in Scaling Biodiversity
(eds Storch, D. & Marquet, P. A. & Brown, J. H.)
(Cambridge University Press, Cambridge, 2006).
71. Martiny, J. B. H. & Field, D. Ecological perspectives on
the sequenced genome collection. Ecol. Lett.
(in the press).
72. Naeem, S., Thompson, L. J., Lawler, S. P., Lawton, J.
H. & Woodfin, R. M. Declining biodiversity can alter
the performance of ecosystems. Nature 368, 734–
737 (1994).
73. Hooper, D. U. &. Vitousek, P. M. The effects of plant
composition and diversity on ecosystem processes.
Science 277, 1302–1305 (1997).
74. Dukes, J. S. Biodiversity and invasibility in grassland
microcosms. Oecologia 126, 563–568 (2001).
75. Treseder, K. K. & Vitousek, P. M. Potential ecosystemlevel effects of genetic variation among populations
of Metrosideros polymorpha from a soil fertility
gradient in Hawaii. Oecologia 126, 266–275
(2001).
76. McGrady-Steed, J., Harris, P. M. & Morin, P. J.
Biodiversity regulates ecosystem predictability. Nature
390, 162–165 (1997).
77. Bell, T., Newman, J. A., Silverman, B. W., Turner, S. L.
& Lilley, A. K. The contribution of species richness and
composition to bacterial services. Nature 436,
1157–1160 (2005).
78. Naeem, S. & Li, S. B. Biodiversity enhances ecosystem
reliability. Nature 390, 507–509 (1997).
79. van der Heijden, M. G. A. et al. Mycorrhizal fungal
diversity determines plant biodiversity, ecosystem
variability and productivity. Nature 396, 69–72
(1998).
80. Cavigelli, M. A. & Robertson, G. P. The functional
significance of denitrifier community composition in a
terrestrial ecosystem. Ecology 81, 1402–1414 (2000).
81. Horz, H.-P., Barbrook, A., Field, C. B. & Bohannan, B.
J. M. Ammonia-oxidizing bacteria respond to
multifactorial global change. Proc. Natl Acad. Sci.
USA 101, 15136–15141 (2004).
82. Bull, A. T., ed. Microbial Diversity and Bioprospecting.
(ASM Press, Washington DC, 2003).
83. Finlay, B. J., Esteban, G. F., Olmo, J. L. & Tyler, P. A.
Global distribution of free-living microbial species.
Ecography 22, 138–144 (1999).
84. Magurran, A. E. Ecological Diversity and Its
Measurement (Princeton University Press, Princeton,
1988).
85. Chao, A., Chazdon, R. L., Colwell, R. K. & Shen, T. J.
A new statistical approach for assessing similarity of
species composition with incidence and abundance
data. Ecol. Lett. 8, 148–159 (2005).
86. Ricketts, T. H. The matrix matters: effective isolation
in fragmented landscapes. Am. Nat. 158, 87–99
(2001).
87. Mantel, N. The detection of disease clustering and a
generalized regression approach. Cancer Res. 27,
209–220 (1967).
88. Borcard, D., Legendre, P. & Drapeau, P. Partialling out
the spatial component of ecological variation. Ecology
73, 1045–1055 (1992).
89. Smouse, P. E., Long, J. C. & Sokal, R. R. Multiple
regression and correlation extensions of the Mantel
test of matrix correspondence. Syst. Zool. 35, 627–
632 (1986).
90. Fulthorpe, R. R., Rhodes, A. N. & Tiedje, J. M. High
levels of endemicity of 3-chlorobenzoate-degrading
soil bacteria. Appl. Environ. Microbiol. 64, 1620–
1627 (1998).
91. Schwalbach, M. S. & Fuhrman, J. A. Wide-ranging
abundances of aerobic anoxygenic phototrophic
bacteria in the world ocean revealed by
epifluorescence microscopy and quantitative PCR.
Limnol. Oceanogr. 50, 620–628 (2005).
92. Garcia-Martinez, J. & Rodriguez-Valera, F.
Microdiversity of uncultured marine prokaryotes: the
SAR11 cluster and the marine Archaea of Group I.
Mol. Ecol. 9, 935–948 (2000).
93. Glaeser, J. & Overmann, J. Biogeography,
evolution, and diversity of epibionts in
phototrophic consortia. Appl. Environ. Microbiol.
70, 4821–4830 (2004).
94. Yeager, C. M. et al. Diazotrophic community
structure and function in two successional stages of
biological soil crusts from the Colorado plateau and
Chihuahuan desert. Appl. Environ. Microbiol. 70,
973–983 (2004).
95. Sliwinski, M. K. & Goodman, R. M. Spatial
heterogeneity of Crenarchaeal assemblages within
mesophilic soil ecosystems as revealed by PCR-singlestranded conformation polymorphism profiling. Appl.
Environ. Microbiol. 70, 1811–1820 (2004).
96. Sliwinski, M. K. & Goodman, R. M. Comparison of
Crenarchaeal consortia inhabiting the rhizosphere
of diverse terrestrial plants with those in bulk soil in
native environments. Appl. Environ. Microbiol. 70,
1821–1826 (2004).
97. Riemann, L. & Middelboe, M. Stability of bacterial and
viral community compositions in Danish coastal waters
as depicted by DNA fingerprinting techniques. Aquat.
Microb. Ecol. 27, 219–232 (2002).
112 | FEBRUARY 2006 | VOLUME 4
98. Pinhassi, J. et al. Spatial variability in
bacterioplankton community composition at the
Skagerrak–Kattegat Front. Mar. Ecol. Prog. Ser. 255,
1–13 (2003).
99. Troussellier, M. et al. Bacterial activity and genetic
richness along an estuarine gradient (Rhone River
plume, France). Aquat. Microb. Ecol. 28, 13–24
(2002).
100. Casamayor, E. O. et al. Changes in archaeal, bacterial
and eukaryal assemblages along a salinity gradient by
comparison of genetic fingerprinting methods in a
multipond solar saltern. Environ. Microbiol. 4, 338–
348 (2002).
101. McArthur, J. V., Kovacic, D. A. & Smith, M. H. Genetic
diversity in natural populations of a soil bacterium
across a landscape gradient. Proc. Natl Acad. Sci.
USA 85, 9621–9624 (1988).
102. Buckley, D. H. & Schmidt, T. M. Diversity and
dynamics of microbial communities in soils from
agro-ecosystems. Environ. Microbiol. 5, 441–452
(2003).
103. Staddon, W. J., Trevors, J. T., Duchesne, L. C. &
Colombo, C. A. Soil microbial diversity and community
structure across a climatic gradient in western
Canada. Biodivers. Conserv. 7, 1081–1092 (1998).
104. Franklin, R. B., Taylor, D. R. & Mills, A. L. The
distribution of microbial communities in anaerobic and
aerobic zones of a shallow coastal plain aquifer.
Microb. Ecol. 38, 377–386 (1999).
105. Franklin, R. B. & Mills, A. L. Multi-scale variation
in spatial heterogeneity for microbial community
structure in an eastern Virginia agricultural
field. FEMS Microbiol. Ecol. 44, 335–346
(2003).
106. Martiny, A. C., Jorgensen, T. M., Albrechtsen, H. J.,
Arvin, E. & Molin, S. Long-term succession of structure
and diversity of a biofilm formed in a model drinking
water distribution system. Appl. Environ. Microbiol.
69, 6899–6907 (2003).
107. Franklin, R. B., Blum, L. K., McComb, A. C. & Mills,
A. L. A geostatistical analysis of small-scale spatial
variability in bacterial abundance and community
structure in salt marsh creek bank sediments. FEMS
Microbiol. Ecol. 42, 71–80 (2002).
108. Rohwer, F., Seguritan, V., Azam, F. & Knowlton, N.
Diversity and distribution of coral-associated bacteria.
Mar. Ecol. Prog. Ser. 243, 1–10 (2002).
109. Yannarell, A. C. & Triplett, E. W. Geographic and
environmental sources of variation in lake bacterial
community composition. Appl. Environ. Microbiol. 71,
227–239 (2005).
110. Hewson, I. & Fuhrman, J. A. Richness and diversity of
bacterioplankton species along an estuarine gradient
in Moreton Bay, Australia. Appl. Environ. Microbiol.
70, 3425–3433 (2004).
Acknowledgements
This work was conducted as part of the Patterns in Microbial
Biodiversity Working Group supported by the National Center
for Ecological Analysis and Synthesis, a centre funded by the
National Science Foundation, the University of California at
Santa Barbara and the State of California. We thank
M. Liebold, G. Muyzer, O. Petchey and D. Ward for useful and
lively discussions, and C. van der Gast for comments on the
manuscript. Any opinions, findings and conclusions or recommendations expressed in this study are those of the authors
and do not necessarily reflect the views of the National
Science Foundation.
Competing interests statement
The authors declare no competing financial interests.
FURTHER INFORMATION
Jennifer B. Hughes Martiny’s homepage:
http://www.brown.edu/departments/EEB/hughes
NCEAS: http://www.nceas.ucsb.edu
Access to this interactive links box is free online.
www.nature.com/reviews/micro
© 2006 Nature Publishing Group